Comparative Study of Modeling Wind Speed Data; a Case Study of Maiduguri, Nigeria

Abstract

Wind speed is the most important factor of the wind energy, because of the random nature of wind speed statistical methods are useful in modeling it. The aim of this study is to model wind speed by considering three different probability distributions which are well suited in fitting distributions with varieties of shapes which are widely used in scientific field and often used to model random effect, monitoring environment, wind and rainfall size: are two parameter Lognormal, the two-parameter Weibull and the two-parameter Gamma. These distributions are applied on the wind speed data recorded at the Nigerian Meteorological Agency (NiMET) Office in Maiduguri, at hub height of 10 meters. The period covered by the data is September, 1985 to December, 2011. In order to determine the distribution that best fit the wind data we used the MLE as method of parameter estimation and Kolmogorov-Smirnov test for decision guide. Based on the results the two-parameter Weibull performs the best fit on the wind speed data histogram.